Whatever happened to PGP?
Is PGP still pretty good for identity?
PGP is often thought of as an encryption system, but your private key is a digital signature that can prove who your message comes from, as well as showing that it hasn’t been tampered with.
The reason a Public Key Infrastructure doesn’t look like a widespread identity system is that it needs a web of trust; if somebody you know has signed my PGP key, then you take their word that I am who I say I am. That works well for close groups of friends – or for the corporations and government departments around the world who rely on PKIs based on the commercial PGP offerings or the OpenPGP SDK that’s now available.
That’s where PGP has really made its mark, says Jon Callas (now the CTO of PGP Corporation; and part of the team that shepherded the commercial side of PGP out of the wilderness where Network Associates left it). “We thought it would be a grass roots system.
But now it’s really corporations setting up PKIs for their own business reasons. The entities are organisations like BMW or Siemens rather than individuals.” Some of the companies who failed to sell PKIs in the past had bad business models that were too expensive: “it was like buying a camera and having to buy 1,000 rolls of film at the same time”.
The OpenPGP standard is one of the things that’s helped PGP become more widespread; the other is the rise of systems that need to identify people on many different platforms. But PGP doesn’t solve all of the problems for that. The Friend of a Friend project (FOAF) uses digital signatures to attach PGP key IDs that verify the email of the address of the author to documents; you still have to decide for yourself if you trust the author which the PGP key identifies. There’s a PKI at the heart of Skype, for example, to make sure you’re talking to the person you want to call. But that tells you nothing about anything else to do with their identity.
PGP software is mature and the technology is both tested and flexible; you can use an LDAP server as a PGP keyserver or use the PGPticket protocol to issue secure authorisations instead of vulnerable passwords for access to a network service.
What many people want to do with identity now means making those identities work more widely. That comes down to the architecture of the systems that will accept identities, and the ways those identities are secured will include PGP (read more about PGP Identity Management here).
Identity management is changing into claims management and different claims will come from different systems, bringing together claims like your Skype ID, your age and your eBay ranking only when someone needs to know you’re old enough to buy what they’re selling before they call you. Some of those claims will be secured and verified by PGP.
Instead of building up as much information about your users as possible – something marketing departments are happier about than the users themselves – you can think about the smallest pieces of information you need for a specific authorisation or transaction.
Whether it’s issues of liability or commercial advantage, businesses don’t want to share more of their customer database than they have to, Callas points out. Ironically, their commercial interests are turning out to enable our desire for privacy. ®
Sponsored: 2016 Cyberthreat defense report